The 3rd-Generation Partnership Project (3GPP) has developed a third-generation wireless communications known as Long Term Evolution (LTE) technology, as documented in the specifications for the Evolved Universal Terrestrial Radio Access Network (UTRAN). LTE is a mobile broadband wireless communication technology in which transmissions from base stations (referred to as eNodeBs or eNBs in 3GPP documentation) to mobile stations (referred to as user equipment, or UEs, in 3GPP documentation) are sent using orthogonal frequency division multiplexing (OFDM). OFDM splits the transmitted signal into multiple parallel sub-carriers in frequency.
More specifically, LTE uses OFDM in the downlink and Discrete Fourier Transform (DFT)-spread OFDM in the uplink. The basic LTE downlink physical resource can be viewed as a time-frequency resource grid. FIG. 1 illustrates a portion of the available spectrum of an exemplary OFDM time-frequency resource grid 50 for LTE. Generally speaking, the time-frequency resource grid 50 is divided into one millisecond subframes. As seen in FIGS. 1 and 2, each subframe includes a number of OFDM symbols. For a normal cyclic prefix (CP) length, which is suitable for use in situations where multipath dispersion is not expected to be extremely severe, a subframe consists of fourteen OFDM symbols. A subframe has only twelve OFDM symbols if an extended cyclic prefix is used. In the frequency domain, the physical resources are divided into adjacent subcarriers with a spacing of 15 kHz. The number of subcarriers varies according to the allocated system bandwidth. The smallest element of the time-frequency resource grid 50 is a resource element. A resource element consists of one OFDM subcarrier during one OFDM symbol interval.
LTE resource elements are grouped into resource blocks (RBs), which in its most common configuration consists of 12 subcarriers and 7 OFDM symbols (one slot). Thus, a RB typically consists of 84 REs. The two RBs occupying the same set of 12 subcarriers in a given radio subframe (two slots) are referred to as an RB pair, which includes 168 resource elements if a normal CP is used. Thus, an LTE radio subframe is composed of multiple RB pairs in frequency with the number of RB pairs determining the bandwidth of the signal. In the time domain, LTE downlink transmissions are organized into radio frames of 10 ms, each radio frame consisting of ten equally-sized subframes of length Tsubframe=1 ms.
The signal transmitted by an eNB to one or more UEs may be transmitted from multiple antennas. Likewise, the signal may be received at a UE that has multiple antennas. The radio channel between the eNB distorts the signals transmitted from the multiple antenna ports. To successfully demodulate downlink transmissions, the UE relies on reference symbols that are transmitted on the downlink. Several of these reference symbols are illustrated in the resource grid 50 shown in FIG. 2. These reference symbols and their position in the time-frequency resource grid are known to the UE and hence can be used to determine channel estimates by measuring the effect of the radio channel on these symbols.
Several techniques may be used to take advantage of the availability of multiple transmit and/or receive antennas. Some of these are referred to as Multiple-Input Multiple-Output (MIMO) transmission techniques. One example technique used when multiple transmit antennas are available is called “transmit precoding,” and involves the directional transmission of signal energy towards a particular receiving UE. With this approach, the signal targeted to a particular UE is simultaneously transmitted over each of several antennas, but with individual amplitude and/or phase weights applied to the signal at each transmit antenna element. This application of weights to the signal is referred to as “precoding,” and the antenna weights for a particular transmission can mathematically be described in a comprehensive way by a precoding vector.
This technique is sometimes referred to as UE-specific precoding. The reference symbols accompanying a precoded transmission and used for its demodulation are denoted a UE-specific reference signal (UE-specific RS). If the transmitted symbols making up a UE-specific RS in a given RB are precoded with the same UE-specific precoding as the data carried in that RB (where data in this sense can be control information), then the transmission of the UE-specific RS and data can treated as though they were performed using a single virtual antenna, i.e. a single antenna port. The targeted UE performs channel estimation using the UE-specific RS and uses the resulting channel estimate as a reference for demodulating the data in the RB.
UE-specific RS are transmitted only when data is transmitted to a UE in the RB pair, and are otherwise not present. In Releases 8, 9, and 10 of the LTE specifications, UE-specific reference signals are included as part of each of the RBs that are allocated to a UE for demodulation of the physical downlink shared data channel (PDSCH). Release 10 of the LTE specifications also supports spatial multiplexing of the downlink transmission, allowing up to eight spatially multiplexed “layers” to be transmitted simultaneously. Accordingly, there are eight orthogonal UE-specific RS, which are described in the 3GPP document “Evolved Universal Terrestrial Radio Access (E-UTRA); Physical Channels and Modulation,” 3GPP TS 36.211, v. 10.0.0 (December 2010), available at www.3gpp.org. These are denoted as antenna ports 7-15. FIG. 3 shows an example of the mapping of UE-specific reference symbols to a RB pair; in this example antenna ports 7 and 9 are shown. Antenna ports 8 and 10 can be obtained as code-division multiplexed reference signals on top of antenna ports 7 and 9, respectively.
Another type of reference symbols are those that can be used by all UEs. These reference symbols must therefore have wide cell area coverage and are thus not precoded towards any particular UE. One example is the common reference symbols (CRS) used by UEs for various purposes, including channel estimation and mobility measurements. These CRS are defined so that they occupy certain pre-defined REs within all the subframes in the system bandwidth, irrespectively of whether there is any data being sent to users in a subframe or not. These CRS are shown as “reference symbols” in FIG. 2.
Another type of reference symbol is the channel state information RS (CSI-RS), introduced in Release 10 of the LTE specifications. CSI-RS are used for measurements associated with precoding matrix and transmission rank selection for transmission modes that use the UE-specific RS discussed above. The CSI-RS are also UE-specifically configured. Yet another type of RS is the positioning RS (PRS), which was introduced in LTE Release 9 to improve positioning of UEs in a network.
Messages transmitted over the radio link to users can be broadly classified as control messages or data messages. Control messages are used to facilitate the proper operation of the system as well as proper operation of each UE within the system. Control messages include commands to control functions such as the transmitted power from a UE, signaling of RBs within which the data is to be received by the UE or transmitted from the UE, and so on.
Specific allocations of time-frequency resources in the LTE signal to system functions are referred to as physical channels. For example, the physical downlink control channel (PDCCH) is a physical channel used to carry scheduling information and power control messages. The physical HARQ indicator channel (PHICH) carries ACK/NACK in response to a previous uplink transmission, and the physical broadcast channel (PBCH) carries system information. The primary and secondary synchronization signals (PSS/SSS) can also be seen as control signals, and have fixed locations and periodicity in time and frequency so that UEs that initially access the network can find them and synchronize. Similarly, the PBCH has a fixed location relative to the primary and secondary synchronization signals (PSS/SSS). The UE can thus receive the system information transmitted in BCH and use that system information to locate and demodulate/decode the PDCCH, which carries control information specific to the UE.
As of Release 10 of the LTE specifications, all control messages to UEs are demodulated using channel estimates derived from the common reference signals (CRS). This allows the control messages to have a cell-wide coverage, to reach all UEs in the cell without the eNB having any particular knowledge about the UEs' positions. Exceptions to this general approach are the PSS and SSS, which are stand-alone signals and do not require reception of CRS before demodulation. The first one to four OFDM symbols of the subframe are reserved to carry such control information; one OFDM symbol is used for this purpose in the example subframe shown in FIG. 2, where the control region may contain up to three OFDM symbols for control signaling. The actual number of OFDM symbols reserved to the control region may vary, depending on the configuration of a particular cell.
Control messages can be categorized into messages that need to be sent only to one UE (UE-specific control) and those that need to be sent to all UEs or some subset of UEs numbering more than one (common control) within the cell being covered by the eNB. Messages of the first type (UE-specific control messages) are typically sent using the PDCCH, using the control region. It should be noted that in future LTE releases there will be new carrier types that may not have such a control region, i.e., that do not have PDCCH transmissions. These new carrier types may not even include CRS, and are therefore not backward compatible. A new carrier of this type is introduced in Release 11. However, this new carrier type is used only in a carrier aggregation scenario, and is always aggregated with a legacy (backward-compatible) carrier type. In future releases of LTE it may also be possible to have stand-alone carriers that do not have a control region and that are not associated with a legacy carrier.
Control messages of PDCCH type are transmitted in association with CRS, which are used by the receiving mobile terminals to demodulate the control message. Each PDCCH is transmitting using resource elements grouped into units called control channel elements (CCEs) where each CCE contains 36 REs. A single PDCCH message may use more than one CCE; in particular a given PDCCH message may have an aggregation level (AL) of 1, 2, 4 or 8 CCEs. This allows for link adaptation of the control message. Each CCE is mapped to 9 resource element groups (REGs) consisting of 4 RE each. The REGs for a given CCE are distributed over the system bandwidth to provide frequency diversity for a CCE. This is illustrated in FIG. 4. Hence, a PDCCH message can consist of up to 8 CCEs spanning the entire system bandwidth in the first one to four OFDM symbols, depending on the configuration.
Processing of a PDCCH message in an eNB begins with channel coding, scrambling, modulation, and interleaving of the control information. The modulated symbols are then mapped to the resource elements in the control region. As mentioned above, control channel elements (CCE) have been defined, where each CCE maps to 36 resource elements. By choosing the aggregation level, link-adaptation of the PDCCH is obtained. In total there are NCCE CCEs available for all the PDCCH to be transmitted in the subframe; the number NCCE may vary from subframe to subframe, depending on the number of control symbols n and the number of configured PHICH resources.
Since NCCE can vary from subframe to subframe, the receiving terminal must blindly determine the position of the CCEs for a particular PDCCH as well as the number of CCEs used for the PDCCH. With no constraints, this could be a computationally intensive decoding task. Therefore, some restrictions on the number of possible blind decodings a terminal needs to attempt have been introduced, as of Release 8 of the LTE specifications. One constraint is that the CCEs are numbered and CCE aggregation levels of size K can only start on CCE numbers evenly divisible by K. This is shown in FIG. 5, which illustrates CCE aggregation for aggregation levels AL-1, AL-2, AL-4, and AL-8. For example, an AL-8 PDCCH message, made up of eight CCEs, can only begin on CCEs numbered 0, 8, 16, and so on.
A terminal must blindly decode and search for a valid PDCCH over a set of CCEs referred to as the UE's search space. This is the set of CCEs that a terminal should monitor for scheduling assignments or other control information, for a given AL. An example search space is illustrated in FIG. 6, which illustrates the search space a particular terminal needs to monitor. Note that different CCEs must be monitored for each AL. In total there are NCCE=15 CCEs in this example. A common search space, which must be monitored by all mobile terminals, is marked with diagonal stripes, while a UE-specific search is shaded.
In each subframe and for each AL, a terminal will attempt to decode all of the candidate PDCCHs that can be formed from the CCEs in its search space. If the Cyclic Redundancy Check (CRC) for the attempted decoding checks out, then the contents of the candidate PDCCH are assumed to be valid for the terminal, and the terminal further processes the received information. Note that two or more terminals may have overlapping search spaces, in which case the network may have to select only one of them for scheduling of the control channel. When this happens, the non-scheduled terminal is said to be blocked. The search spaces for a UE vary pseudo-randomly from subframe to subframe to reduce this blocking probability.
As suggested by FIG. 6, the search space is further divided into a common part and a terminal-specific (or UE-specific) part. In the common search space, PDCCH containing information to all or a group of terminals is transmitted (paging, system information, etc.). If carrier aggregation is used, a terminal will find the common search space present on the primary component carrier (PCC) only. The common search space is restricted to aggregation levels 4 and 8, to give sufficient channel code protection for all terminals in the cell. Note that since it is a broadcast channel, link adaptation cannot be used. The m8 and m4 first PDCCH (where the “first” PDCCH is the one having the lowest CCE number) in an AL of 8 or 4, respectively, belong to the common search space. For efficient use of the CCEs in the system, the remaining search space is terminal specific at each aggregation level.
A CCE consists of 36 QPSK modulated symbols that map to 36 REs that are unique to the given CCE. Hence, knowing the CCE means that the REs are also known automatically. To maximize the diversity and interference randomization, interleaving is used before a cell-specific cyclic shift and mapping to REs. Note that in most cases some CCEs are empty, due to PDCCH location restrictions within terminal search spaces and aggregation levels. The empty CCEs are included in the interleaving process and mapped to REs as any other PDCCH, to maintain the search space structure. Empty CCEs are set to zero power, meaning that the power that would have otherwise been used may be allocated instead to non-empty CCEs, to further enhance the PDCCH transmission.
To facilitate the use of 4-antenna transmit diversity, each group of four adjacent QPSK symbols in a CCE is mapped to four adjacent REs, denoted a RE group (REG). Hence, the CCE interleaving is quadruplex-(group of 4) based. The mapping process has a granularity of 1 REG, and one CCE corresponds to nine REGs (36 RE).
Transmission of the physical downlink shared data channel (PDSCH) to UEs uses those REs in a RB pair that are not used either for control messages (i.e, in the data region of FIG. 4) or RS. The PDSCH can be transmitted using either UE-specific RS or the CRS as a demodulation reference, depending on the PDSCH transmission mode. The use of UE-specific RS allows a multi-antenna base station to optimize the transmission using pre-coding of both data and reference signals being transmitted from the multiple antennas so that the received signal energy increases at the UE and consequently, the channel estimation performance is improved and the data rate of the transmission could be increased.
For Release 11 of the LTE specifications, it has been agreed to introduce UE-specific transmission of control information in the form of enhanced control channels. This is done by allowing the transmission of control messages to a UE where the transmissions are placed in the data region of the LTE subframe and are based on UE-specific reference signals. Depending on the type of control message, the enhanced control channels formed in this manner are referred to as the enhanced PDCCH (ePDCCH), enhanced PHICH (ePHICH), and so on.
For the enhanced control channel in Release 11, it has been further agreed to use antenna port pε{107, 108, 109, 110} for demodulation, which correspond with respect to reference symbol positions and set of sequences to antenna ports pε{7, 8, 9, 10}, i.e., the same antenna ports that are used for data transmissions on the Physical Data Shared Channel (PDSCH), using UE-specific RS. This enhancement means that the precoding gains already available for data transmissions can be achieved for the control channels as well. Another benefit is that different physical RB pairs (PRB pairs) for enhanced control channels can be allocated to different cells or to different transmission points within a cell. This can be seen in FIG. 7, which illustrates ten RB pairs, three of which are allocated to three separate ePDCCH regions comprising one PRB pair each. Note that the remaining RB pairs can be used for PDSCH transmissions. The ability to allocate different PRB pairs to different cells or different transmission points facilitates inter-cell or inter-point interference coordination for control channels. This is especially useful for heterogeneous network scenarios, as will be discussed below.
The same enhanced control region can be used simultaneously by different transmission points within a cell or by transmission points belonging to different cells, when those points are not highly interfering with one another. A typical case is the shared cell scenario, an example of which is illustrated in FIG. 8. In this case, a macro cell 62 contains several lower power pico nodes A, B, and C within its coverage area 68, the pico nodes A, B, C having (or being associated with) the same synchronization signal/cell ID. In pico nodes which are geographically separated, as is the case with pico nodes B and C in FIG. 8, the same enhanced control region, i.e., the same PRBs used for the ePDCCH, can be re-used. With this approach, the total control channel capacity in the shared cell will increase, since a given PRB resource is re-used, potentially multiple times, in different parts of the cell. This ensures that area splitting gains are obtained. An example is shown in FIG. 9, which shows that pico nodes B and C share the enhanced control region whereas A, due to its proximity to both B and C, is at risk of interfering with the other pico nodes and is therefore assigned an enhanced control region which is non-overlapping. Interference coordination between pico nodes A and B, or equivalently transmission points A and B, within a shared cell is thereby achieved. Note that in some cases, a UE may need to receive part of the control channel signaling from the macro cell and the other part of the control signaling from the nearby Pico cell.
This area splitting and control channel frequency coordination is not possible with the PDCCH, since the PDCCH spans the whole bandwidth. Further, the PDCCH does not provide possibility to use UE-specific precoding since it relies on the use of CRS for demodulation.
FIG. 10 shows an ePDCCH that is divided into multiple groups and mapped to one of the enhanced control regions. This represents a “localized” transmission of the ePDCCH, since all of the groups making up the ePDCCH message are grouped together in frequency. Note that these multiple groups are similar to the CCEs in the PDCCH, but are not necessarily made up of the same numbers of REs. Also note that, as seen in FIG. 10, the enhanced control region does not start at OFDM symbol zero. This is to accommodate the simultaneous transmission of a PDCCH in the subframe. However, as was mentioned above, there may be carrier types in future LTE releases that do not have a PDCCH at all, in which case the enhanced control region could start from OFDM symbol zero within the subframe.
While the localized transmission of ePDCCH illustrated in FIG. 10 enables UE-specific precoding, which is an advantage over the conventional PDCCH, in some cases it may be useful to be able to transmit an enhanced control channel in a broadcasted, wide area coverage fashion. This is particularly useful if the eNB does not have reliable information to perform precoding towards a certain UE, in which case a wide area coverage transmission may be more robust. Another case where distributed transmission may be useful is when the particular control message is intended for more than one UE, since in this case UE-specific precoding cannot be used. This is the general approach taken for transmission of the common control information using PDCCH (i.e. in the common search space (CSS)).
Accordingly, a distributed transmission over enhanced control regions can be used, instead of the localized transmission shown in FIG. 10. An example of distributed transmission of the ePDCCH is shown in FIG. 11, where the four parts belonging to the same ePDCCH are distributed among the enhanced control regions.
3GPP has agreed that both localized and distributed transmission of an ePDCCH should be supported, these two approaches corresponding generally to FIGS. 10 and 11, respectively. When distributed transmission is used, then it is also beneficial if antenna diversity can be achieved to maximize the diversity order of an ePDCCH message. On the other hand, sometimes only wideband channel quality and wideband precoding information are available at the eNB, in which case it could be useful to perform a distributed transmission but with UE-specific, wideband, precoding.
Several problems relate to the use of the ePDCCH. For example, if an ePDCCH based on distributed transmission is mapped to all PRB pairs that have been configured for the UE, then it is currently a problem that unused resources in these PRB pairs cannot be simultaneously used for PDSCH transmission. As a result, a large control channel overhead will occur in the event that the fraction of unused resources is large. Another unsolved problem is how to handle the collisions between enhanced control channels and the legacy reference signals such as CSI-RS, CRS, PRS, PSS, SSS and legacy control channels such as PDCCH, PHICH, PCFICH and PBCH.
More generally, remaining challenges include how to design the search space for ePDCCH reception in an efficient manner, so that both the localized and distributed (or UE-specific precoding and diversity transmission) ePDCCH can be supported flexibly for different ePDCCH transmissions.